**Methods** <br> In the pre-registered methods for follow-up study #1, it was implied that subjects were instructed to ignore the first array of objects in the informed version of the change detection task. In fact, they were instructed that if they saw a familiar object in the second array, it was the object that changed and should be selected as their response. Thus they could have completely ignored the first array of objects, but were not directly instructed to. <br> <br> In the pre-registered methods for the main experiement and follow-up study #1, a parameter in the methods for the studied change detection trials was not updated (left as 'tbd'). This value was 12.6 degrees, consistent with the other change detection trials. **Analysis** <br> The pre-registered versions of the analysis scripts calculated effect sizes and confidence intervals using the bootES R library. This method was incorrect, as bootES cannot perform the calculations for within-subjects designs. The final versions of the scripts use a custom function to perform within-subjects effect size and confidence interval calculations.